Impact of Radiation Dosimetry and Clinical Factors on Implant-Based Breast Reconstruction Outcomes Following Post-Mastectomy Radiotherapy

Abstract

<h3>Purpose/Objective(s)</h3> Post-mastectomy radiotherapy (PMRT) is associated with increased risk of implant-based breast reconstruction complications. We sought to examine the relationship between clinical/dosimetric factors and implant reconstruction complications. <h3>Materials/Methods</h3> We conducted a retrospective study of breast cancer patients (pts) who underwent simulation for PMRT from 2012 to 2015 following single-stage or two-stage implant-based reconstruction. Demographic, disease, treatment, and reconstruction outcomes data were collected. Major reconstruction complication (MRC) was defined as any complication requiring unplanned reoperation or removal. Clinical/dosimetric factors associated with MRC were determined by Cox regression. Cutpoints for dosimetric variables were determined by the Contal and O'Quigley method. <h3>Results</h3> We identified 111 pts who underwent PMRT following implant-based reconstruction. 91% underwent two-stage reconstruction with PMRT to a tissue expander (TE) and 9% underwent single-stage reconstruction with PMRT to the permanent implant (PI). All pts were treated using 3D-CRT with conventional fractionation (1.8-2 Gy/day) for the majority of their treatment. 11 pts (10%) received hypofractionated (2.1-3 Gy) treatment for the last 2-5 fractions for scheduling reasons and were excluded from dosimetric analysis. 44% received a chest wall boost (4-10 Gy). At a median follow-up of 6.6 years, 45 pts (41%) experienced an MRC (8 had MRC while TE was in place, 7%; 37 had MRC while PI was in place, 33%). Following MRC, at time of last follow-up/death, 15 pts lacked replacement reconstruction and 12 had autologous reconstruction. Median time from end of PMRT to MRC was 1.6 yrs. The most common types of MRC were contracture (31%), infection (22%), asymmetry (18%), and threatened exposure of the PI (7%). On univariate analysis, there was no significant difference between pts with or without MRC with respect to age, race, smoking history, menopause status, chemotherapy or hormonal therapy, or reconstruction timing. Obesity (<i>P</i> = 0.03; HR 2.26, 1.08-4.76) and infection occurring after mastectomy and within 30 days after PMRT (<i>P</i> < 0.0001; HR 7.37, 2.78-19.55) were significantly associated with MRC. Dosimetric characteristics (excluding boost), including maximum chest wall hot spot ≥ 5464 cGy (P = 0.003; HR 4.08, 1.60-10.41), D_2cc ≥ 5410 cGy (<i>P</i> = 0.01; HR 3.51, 1.37-8.95), D_0.03cc ≥ 5420 cGy (P = 0.01; HR 4.73, 1.46-15.33), V100% ≥ 640 cc (P = 0.01; HR 2.72, 1.26-5.91), V105% ≥ 250 cc (P = 0.003; 2.58, 1.39-4.77), and V108% ≥ 3.5 cc (P = 0.01; HR 2.97, 1.25-7.07) were significantly associated with MRC. <h3>Conclusion</h3> Patients undergoing PMRT and implant-based reconstruction are at significant risk of MRC. Obesity, infection within 30 days of PMRT, increased treatment volume, and plan heterogeneity further increase this risk. Additional studies are needed to validate these findings and determine dosimetric constraints or other prophylactic measures that can limit complication risk.